Sometimes it is noticed with time that it's not the "(more) exact (but too long) calculation", that is remembered at the end,

 

but more simple, schematic or short, easy to remember, version, (majority memory). For example :

 

-SensiblerQM Postulate : Measurement of (sub)system disturb the system.

 

(We need 5 symbols in english, and 6 in french, 7 in german)

 

instead of long incomprehensible calculations ?

 

or

 

-Relativity 2+1 : instead of sheet-long matrix products with Lorentz boost, a geometrical view :

 

1/2 angle between both frames of reference space direction [math]\vec{e}_\phi[/math] and [math]\vec{e'}_\phi[/math] is given by : [math]v_\phi/c[/math]

 

So, maybe just the simpler reality is remembered, even if it's not exact, because of demagogical argument ??

so basically you are saying we should abandon accuracy for something more easily remembered.

 

and the large calculations are not incomprehensible to the people who study it (or even just mathematics).

 

i don't see the point. surely accuracy is more useful than it being easily understandable by a layman.